Virtual Recruitment Effects on Matched Residents in Family Medicine: Experiences From Central Pennsylvania.

Abstract

Background and objectives: As a result of the COVID-19 pandemic, interviews during the 2021 US residency match were conducted virtually, a practice again recommended and repeated by many programs in 2022. The impact of virtual interviews on recruitment and match outcomes has recently been of interest, with results showing the virtual format to be mostly well received by applicants due to cost, travel, and scheduling benefits. Few studies have looked at pre/posttransition comparisons of applicant geographic and demographic data. We compared objective match outcomes between in-person and virtual interviews across three residency programs.

Methods: We conducted a retrospective cross-sectional analysis of National Residency Matching Program data between 2015-2022 across three family medicine residency programs. Primary outcomes were fill rate, average rank position, distance from program, and percentage of underrepresented in medicine demographic status for matched applicants. We compared aggregate in-person data (2015-2019) to aggregate virtual data (2020-2022) for each program using χ², Fisher Exact test, or 2-tailed t tests to 95% confidence.

Results: Saint Joseph Hospital in Reading, Pennsylvania, a 3-year community-based university affiliated program, had significantly more unfilled positions during virtual recruitment (P=.0058). Mount Nittany Medical Center in State College, Pennsylvania, a 3-year community based university-affiliated program, had a significant difference in distance of matched residents' current address (P=.048). Virtual interviews were not associated with significant differences in average position on rank list, average distance from permanent address zip code, or percentage of underrepresented in medicine (URiM) demographic status for matched applicants.

Conclusions: The impact of virtual interviewing on unfilled positions and geographic data is likely site specific and generally small, as some programs had significant structural changes. Further research is needed to confirm the generalizability of these results and explore future comparisons of demographic and geographic characteristics of matched applicants pre/posttransition to the virtual format.